[time-nuts] 10 MHz to 32.768 kHz converter

Alex Pummer alex at pcscons.com
Mon Mar 21 18:31:23 UTC 2016


every injection locked oscillator has jitter, since the oscillator 
slowly --depend on the "tank circuit"s Q will return to it's original 
frequency, what will be interrupted by the next injection, therefore the 
oscillator will have a larger phase jump during the interaction with the 
pulse.
"how fare the oscillator could get" is depend on the "tank circuit"-'s 
Q, therefor the jitter in worst case could be a whole period of the 
oscillator, which is   1/100kHz = 10usec
in case the locking pulse does not let the oscillator go to fare away 
the jitter will be less, but the jitter will be always a dependence of 
the Q of the oscillator's "tank circuit  in this case the crystal, also 
the length of the synchronizing pulse -- the energy of the pulse is 
dependon it's length -- will influence the phase of the synchronized 
oscillator.
there was a very good  paper written on injection locking: R. Adler, “A 
study of locking phenomena in oscillators,”Proc. IEEE, vol. 61, no. 10, 
pp. 1380–1385, Oct. 1973

73
KL6UHN
Alex


On 3/21/2016 6:00 AM, Attila Kinali wrote:
> On Sun, 20 Mar 2016 18:26:16 -0000
> "Martyn Smith" <martyn at ptsyst.com> wrote:
>
>> I  have a real time clock calendar chip that requires a 32.768 kHz crystal.
>> I want to feed it with 10 MHz signal instead, so it is synchronised to my
>> main 10 MHz in a frequency standard I am designing.
> Currently, all that has been discussed were digital solutions.
> But what about using an analog approach?
> If you have a 32kHz crystal oscillator, you can injection lock it
> to the 10MHz signal, by dividing the 10MHz down to 128Hz, then use this
> to form short (as in a couple of ns) pulses, which you then couple
> to the crystal using a small (a couple of pF) capacitor.
>
> Given that the crytsal has an accuracy of better than 100ppm, then
> even a very weak coupling at 128Hz should be enough to keep it locked.
> Upper bound on the jitter is 1/128Hz*100ppm=781ps (very simplified
> calculation, but it should be definitly less than 1-2ns)
>
> 			Attila Kinali
>




More information about the Time-nuts_lists.febo.com mailing list